Research On Semi-Automatic Interpretation Of High Resolution Remote Sensing Image For Geographic Condition Monitoring

With the rapid development of high-resolution remote sensing and mapping of information technology,geography Census mapping geographic information field as a new,extremely important work has become the urgent needs of high-resolution remote sensing technology applications.In this paper,combined with geographic Census work is currently carried out in-depth research and analysis of high resolution remote sensing image interpretation process to achieve basic computer automated or semi-automated processing chain,combined with the geographical Census content and surface characteristics of the western region covered,Construction of High resolution Remote sensing image semiautomatic interpretation system in the west,theories and methods of parallel block segmentation and object feature focuses on high-resolution remote sensing image data by a large knowledge base established,targeted to develop fast and efficient resolution semi-automatic interpretation of remote sensing rate tools,technical guidance and provide the basis for the census geographic conditions.Thesis research carried out the following four aspects:(1)High Resolution Remote Sensing Image Characteristics and quantify the spatial domain.Combined Census content and geographical coverage of the western region of surface features,the establishment of roads,buildings,water,vegetation,bare five geographic Census typical terrain classification system,construction of the western semi-automatic remote sensing image interpretation framework for the western semi-automated processing of high resolution remote sensing image interpretation to provide technical guidance,with some technical and theoretical reference value.(2)Analyzed and studied the techniques and methods of high resolution remote sensing image interpretation of semi-automatic rule sets built.In this paper we were selected roads,buildings,water,vegetation,bare five feature high resolution images of samples,interpretation of surface features by studying the five combinations of features,build roads,buildings,water,vegetation,bare ground semi-automatic interpretation of the rules set for the National census geographic feature quickly and efficiently extract and provide the basis for Solutions information.(3)Analyzed and studied the techniques and methods of high resolution remote sensing image interpretation of semi-automatic rule sets built.In this paper we were selected roads,buildings,water,vegetation,bare five feature high resolution images of samples,interpretation of surface features by studying the five combinations of features,build roads,buildings,water,vegetation,bare ground semi-automatic interpretation of the rules set for the National census geographic feature quickly and efficiently extract and provide the basis for Solutions information.(4)In ENVI/IDL platform for the development,design and programming high-resolution remote sensing image interpretation semiautomatic tools for the Census geography.In QuickBird image Shapingba District of Chongqing University Town area of research for the study,with manual sampling 100 checkpoints artificial interpret the results as a reference for the interpretation of the results consistency ratio quantitative indicators to assess the accuracy of the article developed a semi-automatic solution translation software tools to extract information accuracy analysis results.The results showed that: 85% consistency ratio roads,buildings consistency ratio was 83.3%,water consistency ratio was 100%,90%consistency ratio vegetation,bare consistency ratio was 72%.This article will feature remote sensing interpretation knowledge formation characteristics typical feature Census geography features a combination of knowledge,and to establish the appropriate interpretation of the rule set,direct services to the Census geography actual production applications.Research has certain application value,research will also provide the technical basis to carry out post-geographical conditions monitoring.